Apparatus and Method for Parallel Processing

1. A parallel processing apparatus comprising: a memory; a control unit that determines whether one or more threads can access one or more control blocks of a first container that exists in a direction in which at least one of the one or more threads perform a task; a container generating unit that generates a second container that comprises one or more control blocks based on a result of the determination; and a container management unit that connects the one or more control blocks of the first container or the one or more control blocks of the first container and the one or more control block of the second container and control blocks in which the one or more threads perform tasks, in a ring shape, wherein each of the first and the second containers has a pointer that indicates a next container connected to the each of the first and the second containers, and a reuse count, the reuse count being a number of threads that have completed tasks in all of the one or more control blocks of the each of the first and the second containers, and wherein the pointer and the reuse count are parameters.

2. The parallel processing apparatus of claim 1 , wherein the container management unit sets the pointer to connect the one or more control blocks of the first container and the one or more control blocks of the second container in the ring shape.

3. The parallel processing apparatus of claim 2 , wherein a newly generated container is arbitrarily inserted into a position in the ring shape of preexisting containers.

4. The parallel processing apparatus of claim 1 , wherein the control unit compares the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, with a number of threads that are performing the tasks in the one or more control blocks of the first container, and determines whether the one or more threads can access the one or more control blocks of the first container based on a result of the comparison.

5. The parallel processing apparatus of claim 4 , wherein, if the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, is smaller than the number of threads that are performing the tasks in the one or more control blocks of the first container, the control unit determines that the one or more threads cannot access the one or more control blocks of the first container.

6. The parallel processing apparatus of claim 1 , wherein the control unit determines whether the second container needs to be generated based on the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, and progress of the tasks performed by the one or more threads.

7. The parallel processing apparatus of claim 6 , wherein the control unit checks the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, and the progress of the tasks performed by the one or more threads, using a separate thread.

8. The parallel processing apparatus of claim 1 , wherein, if it is determined that the one or more threads cannot access the one or more control blocks of the first container, the container management unit generates the second container.

9. The parallel processing apparatus of claim 1 , wherein the one or more threads comprises a master thread and a plurality of slave threads branched from the master thread.

10. The parallel processing apparatus of claim 9 , further comprising a thread generating unit that generates the master thread and the slave threads.

11. A parallel processing method comprising: determining whether one or more threads can access one or more control blocks of a first container that exists in a direction in which at least one of the one or more threads perform a task; generating a second container that includes one or more control blocks based on a result of the determining; and connecting the one or more control blocks of the first container or the one or more control blocks of the first container and the one or more control blocks of the second container and control blocks in which the one or more threads perform tasks, in a ring shape, wherein each of the first and the second containers has a pointer that indicates a next container connected to the each of the first and the second containers, and a reuse count, the reuse count being a number of threads that have completed tasks in all of the one or more control blocks of the each of the first and the second containers, and wherein the pointer and the reuse count are parameters.

12. The parallel processing method of claim 11 , wherein the connecting in the ring shape comprises setting the pointer to connect the one or more control blocks of the first container and the one or more control blocks of the second container in the ring shape.

13. The parallel processing method of claim 12 , wherein a newly generated container is arbitrarily inserted into a position in the ring shape of preexisting containers.

14. The parallel processing method of claim 11 , wherein the determining comprises comparing the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, with a number of threads that are performing the tasks in the one or more control block of the first container to determine whether the one or more threads can access the one or more control blocks of the first container.

15. The parallel processing method of claim 14 , wherein the determining comprises if the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, is smaller than the number of threads that are performing the tasks in the one or more control blocks of the first container, determining that the one or more threads cannot access the one or more control blocks of the first container.

16. The parallel processing method of claim 11 , further comprising determining whether the second container needs to be generated based on the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, and progress of the tasks performed by the one or more threads.

17. The parallel processing method of claim 16 , wherein the determining whether the second container needs to be generated comprises checking the number of threads that have completed the tasks in all of the one or more control blocks of the first container, which is a parameter of the first container, and the progress of the tasks performed by the one or more threads, using a separate thread.

18. The parallel processing method of claim 11 , wherein the generating the second container comprises if it is determined that the one or more threads cannot access the one or more control blocks of the first container, generating the second container.

19. The parallel processing method of claim 11 , wherein the one or more threads comprise a master thread and a plurality of slave threads branched from the master slave.

20. The parallel processing method of claim 19 , further comprising generating the master thread and the slave threads.